Association of Serum Polyamines with Cardiovascular Events and All-Cause Mortality in Chronic Kidney Disease

Chronic kidney disease (CKD) has become a significant global health problem, causing high morbidity and mortality and posing a major financial burden on healthcare systems globally [1]. As CKD progresses, patients face an elevated risk of developing severe complications, with cardiovascular disease (CVD) being the most prominent. CVD accounts for over half of deaths in advanced CKD and end-stage kidney disease [2, 3]. Recent research has emphasized the critical role of metabolic disorders in the progression of CKD. Polyamines, such as spermidine, spermine, and putrescine, are essential biomolecules in human cell metabolism [4]. Disturbances in polyamine metabolism have been observed in various acute and chronic diseases, including cancers [5‒7], and they hold the potential as diagnostic markers for certain conditions [8]. Putrescine is regarded as a uremic toxin [9], while spermine and spermidine have shown beneficial effects in a range of non-cancerous diseases [6]. These natural polyamines, characterized by their multiple amino groups, can interact with negatively charged macromolecules, influencing processes such as the immune response, chromatin organization, gene regulation, and cell proliferation [10, 11].

Clinical and experimental studies have established links between kidney diseases and changes in polyamines and their regulatory enzymes. An imbalance in kidney polyamines may contribute to different forms of kidney injury [12‒14]. Notably, there is a negative correlation between plasma spermine levels and serum creatinine as well as urea nitrogen [15]. The increased breakdown of spermine in CKD patients may be attributed to enhanced spermine oxidase activity [13].

Numerous animal-based studies have shown that spermine or spermidine play significant roles in anti-fibrotic and anti-aging processes [16, 17], and external supplementation has been found to mitigate the fibrosis in myocardia, liver or lung [18‒21]. However, the addition of putrescine has been reported to exacerbate cardiac hypertrophy in nephrectomy models [22]. Collectively, these studies suggest that serum polyamines may play a role in CKD and its cardiovascular (CV) complications. Nevertheless, it remains unclear whether disturbances in polyamine levels are associated with an increased risk of CV events and all-cause mortality, particularly in non-dialysis CKD patients. Therefore, this study aimed to enroll CKD patients at different stages to investigate the relationship between polyamines, CV morbidity, and mortality, and to evaluate the clinical significance of serum polyamines in non-dialysis CKD patients.

This retrospective cohort study was conducted at Xinqiao Hospital of the Army Medical University in China, involving participants with CKD at stages 1–5 who were not on dialysis yet. Between March 2015 and September 2018, 368 patients were recruited, and they were followed up until May 2023. The serum levels of putrescine, spermidine, and spermine were measured in the enrolled individuals. The study protocol received ethical approval from the Xinqiao Hospital Ethical Committee (No. 2024-269-03).

From March 2015 to September 2018, our study recruited adult non-dialysis CKD patients (stages 1–5) with detailed medical information and follow-up statistics. Exclusion criteria included prolonged usage of drugs that suppress the immune system, acute kidney failure, recent dialysis within the last 6 months, being pregnant, having an infection within the last month, and a history of malignancy. Besides, patients with a sample volume of less than 200 μL or those with hemolysis were excluded as this could affect the detection efficiency of polyamines. To avoid biasing the research conclusions, we also excluded aberrant detection values of polyamines. Ultimately, a total of 297 individuals were included in the analyses (Fig. 1).

At enrollment, clinical data were obtained from Xinqiao Hospital’s Electronic Medical Record System. Blood samples were collected after an overnight fast of at least 8 h and analyzed for standard laboratory parameters. These included serum hemoglobin, albumin, uric acid, calcium, phosphorus, total cholesterol, triglycerides, and intact parathyroid hormone (iPTH). Measurements were carried out using a Beckman AU5800 automatic biochemical analyzer following the manufacturer’s instructions. The estimated glomerular filtration rate (eGFR) was calculated using the creatinine equation from the CKD Epidemiology Collaboration (CKD-EPI) [23] and was used to classify CKD stages according to KDIGO guidelines [24].

Blood samples were collected, centrifuged at 3,000 rpm for 10 min at room temperature, and the sera were stored at −80°C for subsequent analysis. Serum polyamines were quantified using a modified high-performance liquid chromatography method with pre-column derivatization and fluorescence detection [25]. Briefly, 100 μL of serum was mixed with 5% perchloric acid to precipitate proteins, followed by centrifugation. The supernatant was subjected to derivatization by mixing with the internal standard (hexamethylenediamine), sodium carbonate, borate buffered saline, and Dns-Cl reagent. The mixture was incubated in a 60°C water bath for 40 min in the dark. l-proline was then added to consume excess derivatives, and the solution was extracted with chloroform. After centrifugation and nitrogen blow-drying, the residue was dissolved in acetonitrile and analyzed using C18 high-performance liquid chromatography columns (250 × 4.6 mm, 5 μm) with a Waters 2475 Multi-wavelength Fluorescence Detector.

The primary outcomes were CV events and all-cause mortality. CV events were defined as a new onset of CV conditions, including acute myocardial infarction (AMI), unstable angina, stroke, hospitalization due to heart failure, severe arrhythmia, and peripheral artery disease. All-cause mortality referred to death from any cause. Data on CV events were obtained from patients’ medical records.

We summarized participant characteristics using descriptive statistics, presenting means and standard deviations for normally distributed variables, medians for non-normally distributed variables, and percentages for categorical variables. To compare characteristics across serum polyamine quartiles, we employed ANOVA or the Kruskal-Wallis H test for continuous variables and the χ² test for categorical variables. Spearman’s rank correlation, a non-parametric measure of statistical dependence between two variables, was used to examine the correlations between serum polyamines and eGFR, hemoglobin, calcium, phosphorus, iPTH, and other baseline factors.

Serum polyamines were divided into quartiles, with the lowest quartile serving as the reference group. Kaplan-Meier curves and a log-rank test were utilized to evaluate the survival probabilities of CV events and overall mortality based on serum polyamine levels. Cox proportional hazards models were applied to explore the relationships between serum polyamines and CV events and all-cause mortality. The Cox proportional hazards model was used to calculate the adjusted hazard ratio (HR) and 95% confidence interval (CI) for CV events and mortality, adjusting for age, sex, and other variables. The effect of potential confounders that are associated with polyamines was analyzed by constructing models with incremental adjustments as follows: unadjusted (model 1), adjusted for age, sex, body mass index (BMI), and systolic blood pressure (model 2); and model 2 plus diabetes mellitus, current smoking, prevalent CV disease, eGFR, total serum cholesterol, serum triglycerides, and iPTH (model 3). To exclude the impact of death as a competing risk, we utilized the Fine-Gray model to examine the association between the polyamine quantiles and the risk of CV events. Gray’s test was used for statistical difference in competing-risk analysis. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the prediction accuracy for CV events and overall mortality. The area under the receiver operating characteristic curve was compared using the Delong test. The statistical analyses were mainly conducted using IBM SPSS Statistics (version 26) or R software (version 4.4.2). A two-sided p < 0.05 was considered statistically significant for all the analyses.

Initially, 368 CKD patients were screened, and 71 were excluded. Finally, 297 patients were included in the final analysis. The baseline characteristics are presented in Table 1. The mean age of the patients was 46.41 years, and 54.55% were male. The median eGFR was 34.41 mL/min/1.73 m² (interquartile range [IQR] 13.90–56.29). Among the patients, 25.59% were current smokers, and 55.89% had hypertension. The patients were divided into four groups according to CKD stages. As shown in Table 1, compared to patients with eGFR above 60 mL/min/1.73 m², those in lower eGFR groups had higher levels of systolic blood pressure, uric acid, and serum phosphate, but lower levels of serum calcium and hemoglobin. Although the incidence of CV events increased in higher CKD stages, there was no difference in the types of CV events. Then, the serum polyamines levels were measured, and the median serum levels of putrescine, spermidine, and spermine were 6.75 μg/L, 5.85 μg/L, and 11.53 μg/L, respectively. Interestingly, with CKD progression, the serum putrescine levels increased, whereas spermidine and spermine levels decreased significantly (Fig. 2a–c). Moreover, putrescine was negatively correlated with eGFR, while spermidine and spermine were positively correlated with eGFR (Fig. 2d–f).

The association between serum polyamines and clinical parameters were also analyzed. Intriguingly, spermidine and spermine showed strong positive correlations with hemoglobin, serum calcium, and inverse relation with systolic blood pressure, uric acid, phosphorus, and iPTH. In contrast, putrescine exhibited different correlations with these indicators (Table 2). The baseline characteristics of the cohort according to polyamines quartiles are presented in online supplementary Table S1–3 (for all online suppl. material, see https://doi.org/10.1159/000545054). The median putrescine values of the quartiles were 4.28, 6.07, 7.27, and 9.67 μg/L, respectively. Compared with the patients in the lowest putrescine quartile, those in the other three quartiles tended to have higher levels of systolic blood pressure, uric acid, serum phosphorus and iPTH, and lower levels of eGFR, hemoglobin, spermidine and spermine (online suppl. Table S1). However, the higher quartiles of spermidine and spermine showed opposite trends to putrescine in these parameters (online suppl. Table S2–3). To explore the relationship between the underlying etiology of CKD and polyamines, the study population was categorized into three groups, including primary renal diseases, secondary renal diseases, and others. It was indicated that among the polyamines, only spermidine in different quartiles showed significant difference in the etiological classifications.

Over a median follow-up period of 6.11 years (range 3.51–7.62 years), 78 (26.26%) patients suffered CV events. Table 3 shows the hazard ratios of CV events in different polyamine quartiles. Kaplan-Meier survival analysis indicated that the incidence of CV events in patients in the lowest quartile of putrescine was lower than that in the highest quartile (Fig. 3a). Conversely, higher levels of spermidine and spermine were associated with reduced CV event rates (Fig. 3b, c). A Cox proportional hazards regression model was used to analyze the relationship between serum polyamines levels and CV events. After adjusting for age, sex, BMI, systolic blood pressure, diabetes mellitus, current smoking, prevalent CV disease, eGFR, total serum cholesterol, serum triglycerides, and iPTH, higher quartiles of putrescine were associated with a significantly increased risk of CV events compared with the first quartile (Q1). Specifically, participants in the third quartile of putrescine had more than twice the risk of CVD (HR = 3.161, 95% CI 1.148–8.699, p = 0.026), and those in the fourth quartile had an approximately 6-fold greater risk of CV events (HR = 6.972, 95% CI 2.520–19.294, p < 0.001). In contrast to putrescine, spermidine and spermine displayed protective roles in CVD in the fully adjusted model 3. Patients in Q4 of spermidine (HR = 0.077, 95% CI 0.022–0.274, p < 0.001) and spermine (HR = 0.180, 95% CI 0.061–0.530, p = 0.002) had remarkably lower risk of CV events. Given that CV endpoints included the composite of intermediate endpoints in which mortality is a competing event, a Fine-Gray model adjusting for mortality as competing event was performed and the relationship between serum polyamines and CVD remained robust (online suppl. Table S4).

To further explore which types of CV events are associated with polyamines, we classified these events into coronary heart disease, heart failure, stroke, and other outcomes. It was indicated that the number of CV events varied across different polyamine quartiles (online suppl. Table S1–3). As demonstrated in the unadjusted Cox model, higher quartiles of putrescine exhibited a closer association with an increased risk of heart failure and stroke, while higher quartiles of spermidine and spermine showed significantly decreased risk of coronary heart disease and heart failure (online suppl. Table S5). Furthermore, ROC analysis indicated that the reciprocal of spermidine had higher predictive value for CV events than the reciprocal of spermine (p = 0.016) but did not obtain statistical difference in comparison with putrescine (p = 0.288) (Fig. 4a).

During follow-up, 43 (14.47%) CKD patients died. There was a significant association between the serum polyamines and the risk of all-cause mortality. Kaplan-Meier analysis revealed that high levels of putrescine and low levels of spermidine and spermine were associated with an increased risk of mortality (Fig. 3d–f). In addition, a multivariate Cox proportional hazards model was used to further examine the relationship between serum polyamines and all-cause mortality (Table 3). In multivariate Cox regression analyses, the role of putrescine in predicting the risk of mortality appeared to be less conclusive. After the adjustment for age, sex, BMI, and systolic blood pressure, higher quartiles of spermidine and spermine were associated with lower risks of mortality in model 2. However, neither achieved statistical significance in model 3, which adjusted for the crucial confounder eGFR. Subsequent ROC analysis indicated that reciprocal of spermidine had the highest area under the curve for predicting all-cause mortality. Nevertheless, the difference was not statistically significant compared with putrescine and reciprocal of spermine (Fig. 4b).

In this study, we investigated the relationship between serum polyamine levels, CV events, and mortality in 297 patients with CKD. Our findings indicated that high putrescine levels serve as a significant predictor of CV events. Conversely, elevated levels of spermidine and spermine were associated with a reduced risk of adverse clinical outcomes.

The polyamines are derived either from intracellular catabolism of arginine or from dietary sources and the gut microbiome [26]. Since the 1970s, these compounds have been recognized to be associated with CKD and play a crucial role in the pathogenesis of renal disorders [27]. A substantial increase in putrescine levels might be implicated in the pathogenesis of cardiac hypertrophy in animal models [21, 28]. The findings of this study have deepened our understanding of putrescine’s potential as a risk factor for both CV events and the progression of CKD.

Our findings are consistent with previous studies that have validated the beneficial effects of spermine and spermidine supplementation on kidney conditions in disease models. An earlier investigation demonstrated that spermidine could potentially prevent chronic renal interstitial fibrosis by activating the Nrf2 pathway in models of unilateral ureteral obstruction (UUO) [29]. Similarly, exogenous spermine supplementation in UUO mice effectively alleviated renal fibrosis by coordinating autophagy and inhibiting senescence [30]. The beneficial effects of spermine and spermidine can be attributed to their ability to induce autophagy [31], alleviate metabolic endotoxemia, and enhance the intestinal barrier integrity [32]. In the present study, we showed that the levels of spermine and spermidine declined along with CKD progression, which may partially explain the high prevalence of renal fibrosis in CKD patients.

Moreover, the protective roles of spermine and spermidine in CV events were also observed in our research, which was consistent with the prior studies [33, 34]. Based on data from several public database resources, the inverse association between dietary polyamines and CVD mortality was reported in an epidemiological study [33]. Furthermore, in a single-center prospective cohort study of patients with AMI, it was revealed that those with higher serum spermidine levels had a lower risk of recurrent AMI [35]. The following molecular mechanisms may explain the link between spermine, spermidine and cardioprotection. First, the protective effects of spermine and spermidine in cardioprotection are closely concerned with enhancing myocardial autophagy [34]. The administration of spermidine in old mice delays cardiac aging by improving diastolic function and cardiomyocyte structure. Likewise, Yan et al.’s research indicated that the amelioration of cardiac dysfunction after a heart attack can be achieved by increasing autophagy via spermidine-mediated regulation of the AMPK/mTOR signaling pathway [36]. Second, animal studies have shown that spermidine can protect cells from oxidative stress damage by preventing inflammation and the accumulation of free radicals [37]. In rats after a heart attack, spermidine reduced oxidative stress damage by increasing superoxide dismutase levels and decreasing malondialdehyde levels in cardiomyocytes [36]. Collectively, although the precise molecular mechanisms of exogenous spermine and spermidine remain partially unpredictable in different cells under various conditions, some known molecular targets and relevant cellular processes, such as autophagy and translation, can largely maintain the function of cardiomyocytes.

Interestingly, numerous studies have explored the potential anti-aging effects and longevity-promoting properties of spermine and spermidine [34, 38]. The median lifespan of mice supplemented with spermidine or spermine was significantly longer than that of the control group consuming normal drinking water [34]. A decline in tissue spermidine concentrations with aging has been observed in both model organisms and humans, which is attributed to the bioavailability and metabolism of polyamines [39‒41]. Supplementation with spermine or spermidine may offer potential insights into preventing aging-related diseases. Nonetheless, the systemic use of polyamines in patients with kidney disease may be restricted by adverse effects. Careful consideration should be given to the dosage of spermine or spermidine supplementation as low levels can provide protection, while elevated spermine concentrations can be cytotoxic [42]. Additionally, it is crucial to carefully check individuals with kidney disease before recommending spermine or spermidine intake. Based on contemporary research studies, individuals with cancer, specific viral infections, pulmonary hypertension, and asthma are advised against using polyamine as a supplement. Numerous studies have explored the strong connection between polyamines and cancer [43, 44]. In the context of our study, the close association between serum polyamines and the risk of death was clearly demonstrated through both Kaplan-Meier analysis and the crude Cox regression analysis. Spermidine and spermine retained a robust protective effect after adjusting for age, sex, BMI, and systolic blood pressure. However, such a protective role was notably attenuated upon adjusting for a range of potential covariates, especially eGFR. One possible explanation for this outcome could be the relatively limited number of death events in our dataset since the statistical power was inevitably diminished. Additionally, it is quite conceivable that polyamines are intricately involved in multiple signaling pathways and interact with numerous molecules, which might potentially obscure the direct link between polyamines and mortality.

Our study also has several limitations. First, due to the relatively small sample size of our study, which may have an impact on the results, a larger population is needed. Second, we did not measure the level of acrolein, a downstream metabolite of polyamine [13, 45]. Previous reports have suggested serum acrolein level may be a reliable biomarker of chronic renal failure [8]. Last but not least, given that our retrospective cohort study was merely conducted in a single center, it is inevitable to rule out the possibility of selection bias and temporal bias; thus, multi-center prospective studies are required to verify the accuracy of the results in the future.

In conclusion, our study revealed that CKD patients exhibit significant disruption in their serum polyamine levels, which are correlated with kidney function. These altered polyamine levels are associated with an increased risk of CV events and overall mortality. Therefore, serum polyamines show promise as effective prognostic indicators for individuals with CKD.

The study was conducted in accordance with the Declaration of Helsinki and approved by the Xinqiao Hospital Ethical Committee (No. 2024-269-03). All patients included in the analysis provided written informed consent to participate in the study.

The authors have no conflicts of interest to declare.

This research was funded by the Joint Funds of the National Natural Science Foundation of China (Grant No. U22A20279), Key Project of Chongqing Technology Development and Application Program (Grant No. CSTB2023TIAD-KPX0060), Natural Science Foundation Project of Chongqing (Grant No. CSTB2024NSCQ-MSX0768), and Basic Research Program of Natural Science from Science and Technology Department of Shaanxi Province (Project No. 2022JQ-786).

Conceptualization: J.X. and J.Z.; methodology: Z.C. and S.W.; software: Z.C. and L.Y.; validation: L.Y., J.X. and J.Z.; investigation: Z.C., X.X., and S.W.; resources: L.L.; data curation: Z.C. and L.L.; formal analysis, visualization, and writing – original draft preparation: Z.C.; writing – review and editing and project administration: J.X.; supervision: J.Z.; and funding acquisition: X.X. and J.Z. All authors have read and agreed to the published version of the manuscript.

All data generated or analyzed during this study are included in this article and its online supplementary material files. Further inquiries can be directed to the corresponding author.